Wednesday, August 1, 2012

Least Correctional Service

Summary:
The idea of "Least Correctional Service" is to sentence a convicted criminal with the least possible correctional services necessary to prevent recidivism.  As any government spending is "opportunity cost", the money saved can instead be used to deter crime.  Additionally, a convicted criminal will "shift" out of correctional services rather than simply finish a sentence.  In this sentence, if they were originally placed in prison, then they shift to a lighter prison, then a half-way prison, then a curfew and then finally full freedom.

In short, it is hoped that by placing a larger emphasis on analysis of correctional services and their efficacy, the justice system can base decisions more on science than on tradition to lower the overall recidivism rate and cost less money.

Long Description:
Today, the justice system bases punishment on tradition or case precedence.  Adjustments made to the punishment system rests largely on politics, with (in general) the right-side of politics favouring heavier punishments and the left side of politics favouring lighter punishments.  The concept of "Least Correctional Services" implies that rather than favouring heavier or lighter punishments, the decision is based on the efficacy of the sentence.

When a person is convicted of a crime, the severity, the motive, context, background of the individual and anything else that may be relevant are all taken into consideration.  This wealth of information is then used by experts, likely to be judges, in addition to whatever tools may be useful (such as computer software to perform statistical predictions), to craft the "optimal" sentence.

To define the "optimal" sentence, a definition of "success" is necessary for the justice system.  The definition used in this article is "reach the lowest possible recidivism rate given the current budget".

Correctional services has two main aspects in the LCS model.

  • Be immediately placed in the least restrictive correctional service possible
  • Convicted shift between more and less restrictive correctional services
As an example, let us pretend there is a person A who is convicted of assault and theft.  It is determined that A was attempting to steal money from victim B to pay for a drug addiction.  According to statistical analysis, past response to such criminal activity, that the LCS model determines this person be placed into drug rehabilitation and curfew, only allowed out to work at a day job.

Upon a 6-month review, it is determined that A is responding poorly to drug rehabilitation and has trouble finding a job.  Therefore, it is determined they enter a more invasive drug rehabilitation program.  Or perhaps, upon a 6-month review, it is determined that A is responding well to drug rehabilitation therapy and therefore is scheduled for restrictions to be softened by the next review.

A large proportion of the lowest levels, such as rehab, community work, curfew relies on a job placement program.  Any such job placement program should also be available publicly, at large, so that everyone can be given aid in finding a position that matches their skill level.  A lack of jobs in an economic downturn is a different topic.

Under this system the most powerful punishment is farm prison and the least is an enforced curfew.  Breaches of conditions results in worsening conditions.  Good reviews result in lifting of restrictions.  Farm prisons, rather than other types of prisons is the worst punishment possible to avoid prison gang membership and criminals training criminals in prison.

At first this system will have broad recommendations of punishment for "classes" of convicted criminals.  For instance, all people who match broad characteristics will receive 16 months of farm prison.  The ideal goal is to have individualised correctional plans.


Rationale:
The idea of LCS is to remove politics and emotions from the justice system and instead give it a clear goal for society.  In this way it can efficiently spend resources to reduce crime and create a higher quality of life in a country.  Pointlessly sentencing people to long sentences, or being "tough on crime" has had no statistical effect on crime nor any perceivable difference in deterrence.

However, it is known that these tough on crime policies are exceptionally costly to the point where justice systems fail.  It can be seen in British Columbia where thousands are now being released because there is insufficient funding, or in California where the same has occurred.  Even a state, such as Texas, has turned away from jailing people for long periods of time simply because they have run out of money.

It is not sustainable to maintain a justice system that continues to spend so many resources to achieve so little.

Tools:
The LCS model requires a lot of data.  The most likely candidate to collect this information is statcan, in an anonymous aggregated manner, to maintain the integrity of the data.

In essence, it is trying to compile data with respect to recidivism rate:

  • How does recidivism rate change with type of crime?
  • How does recidivism rate change with motive of perpetrator?
  • How does recidivism rate change with level of punishment?
As we determine more relevant characteristics then StatCan will be given a mandate to track those additional pieces of information.  The cost of maintaining such data is hopefully offset by the decrease in cost of the justice system.

In additional, an expansion of the job placement program in Canada should be made so that it is more accessible and can be used by anyone, rather than simply suggested to people on Employment Insurance.  Generally this would be a job board where all employers post up jobs, and people have a one-click button to apply (and send a resume/CV to an employer to review).